MicroSquirt.com read only forum (Jan.15,2018)

Hi,
Ive been working on these and from testing this weekend they work quite well. I added two additional temperature probes by duplicating the clt/mat circuit. These were placed pre and post turbo which allows me to calculate the adiabatic efficiency of the turbo.
I can also calculate the efficiency of the intercooler using the turbo outlet and MAT temperatures.

Here's the maths required for megatune:

; Get raw values from ADC.
x6ADC = scalar, U08, 26, "ADC", 1.000, 0.000 ;x6 turbo inlet probe (or ambient temp)
x7ADC = scalar, U08, 27, "ADC", 1.000, 0.000 ;x7 turbo outlet probe

Then do some maths:

;convert to deg 'f for calcs
tinletF = { table(x6ADC, "matfactor.inc")-40} ; Turbo Inlet (atmosphere) deg 'f
toutletF = { table(x7ADC, "matfactor.inc")-40} ; Turbo outlet deg 'f

;Convert into user units for guages
auxmat = { tempCvt(table(x6ADC, "thermfactor.inc")-40) } ; Manifold temperature in user units.
auxmat1 = { tempCvt(table(x7ADC, "thermfactor.inc")-40) } ; Manifold temperature in user units.

;Calculate Turbo Adiabatic efficiency
P1 = { barometer * 0.1450377 } ; inlet pressure absolute (barometer in psi) should be ~14.7
T1 = { tinletF + 460 } ;inlet temp in Radians
P2 = { P1 + boost } ;outlet pressure absolute
T2 = { (T1 * ( pow (P2 / P1, 0.283)) -460)} ; temperature rise assuming 100% compressor Efficiency
AE = { ((T2 - tinletF)/(toutletF - tinletF)) * 100 } ; Adiabatic Efficiency % is difference between actual and calculated temp rises.
;Calculate Intercooler efficiency
;Static ambient temp used here, because my turbo inlet temp isn't the same as ambient due to heat wash over filter.
ambient = { 10 } ; in user units.
ICE = { ((auxmat1 - mat) / (auxmat1 - ambient) * 100) } ;Intercooler/Chargecooler efficiency in %

Then add these to your Gauges section
;Turbo Gauges
auxmatGauge = auxmat, "Turbo Inlet", "%TEMP", -40, 110, -15, 0, 95, 100, 3, 0
auxmatGauge1 = auxmat1, "Turbo Outlet", "%TEMP", -40, 110, -15, 0, 95, 100, 0, 0
turbo = AE, "Adiabatic Efficiency ", "%", 0, 100, -1, -1, 100, 100, 1, 1
intercooler = ICE, "Intercooler Efficiency ", "%", 0, 100, -1, -1, 100, 100, 1, 1


And of course you can log the data and draw pretty graphs.

You can see once the temperature settles down, I'm seeing about 65-70% turbo adiabatic eff and only 50% chargecooler eff.

More details are here:
http://www.mez.co.uk/turbo10.html

"eliotmansfield" wrote:
> You can see once the temperature settles down, I'm seeing about 65-70% turbo adiabatic eff and only 50% chargecooler eff.

I suspect you'll find that the efficiency of the chargecooler improves
substantially when you really start pushing it. As it stands the temperature
differentials are pretty small, start getting your boost cranked up so
your turbo outlet temps are 100+degC and you'll see the chargecooler making
a lot more difference.
Posted by email.

Hey you could take it a step further and add these fields your datalog with these entries in the [Datalog] section of the ini:

entry = auxmat, "Turbo Inlet", int, "%d"
entry = auxmat1, "Turbo Outlet", int, "%d"
entry = AE, "Adiabatic Efficiency", int, "%d"
entry = ICE, "Intercooler Efficiency", int, "%d"

LT401Vette wrote:Hey you could take it a step further and add these fields your datalog with these entries in the [Datalog] section of the ini:

entry = auxmat, "Turbo Inlet", int, "%d"
entry = auxmat1, "Turbo Outlet", int, "%d"
entry = AE, "Adiabatic Efficiency", int, "%d"
entry = ICE, "Intercooler Efficiency", int, "%d"

Yes, already doing that. That's how i generated the graph above.

I must be looking at this wrong. Am I the only one that sees the turbo compressor efficiency go above 100%?

You need to allow for the response time of the thermistor, plus the heat sink effect of the associated plumbing including the compressor housing.

It goes above 100% because the presure rise is instant, whilst the temperature rise is gradual.
So at that instant moment where you have boost, but no [apparent] temperature rise, it appears to be over efficient. Watch as the temperature rises, the efficiency comes down to more sensible values.

So I need to add some hysteresis functions to the expression language to allow you to account for the lag?

Eric

Not for me. I just wasn't thinking of the lag.

Pay no attention to the man behind the curtain!

Nice job, by the way.

Wow, awesome data. I think I'll set this up for my setup so I can test some different things. Thanks for the info!

Good job!! -- I was wanting to do exactly this when I finally get my turbo project revived again. I'll bookmark this for when that happens.